There have been some studies with long term observational data to verify the findings of randomized controlled trials. Observational data typically involve a population that is not as selected with the criteria used in a trial and thus may show different results. I have not done any long-term cohort studies myself but my colleagues at the University of Alabama at Birmingham (UAB) have done several studies examining the use of various treatments:
Following people who participated in a trial for many years after the trial ends presents the practical problem that medical technology changes so that participants may switch from what they received in the trial to something that may be newer, cheaper, more expensive, or maybe more effective.
In addition to study length vs treatment duration: the patient in front of you has many variables ( age, coexisting conditions, concomitant drugs, differing lifestyles etc) from study participants that further complicate the “decision to treat “.
This clinical question is pervasive and implicates multiple medications in most patients (at least in cardiology, but I suspect in any field that deals with chronic health conditions). I saw many patients today where this would be a conundrum for many meds.
Shared decision making- let your patients know where evidence ends (as in extending the findings of clinical trials beyond their trial duration), and where (hopefully) expert opinion begins. And let them choose what to do. I have been taking a statin for > 20 years for primary prevention- there is no evidence that I can quote to support this practice. But, I know that my father had severe CAD, and had a 5 vessel CABG at a much earlier age. So I live with the uncertainty of whether long-term statins for primary prevention is good or not. I choose to believe (right or wrong) that clinical benefits continue to accumulate over time (in the case of statins to lower LDL), and thus I choose to continue to take them.
The lack of long-term followup is one large problem; thanks for the post. A related problem is that effectiveness in the real world is normally less than in the trial, for many reasons. For both of these, CMS could require pharma and device makers to carry out long-term observational studies, and should. No one else can, I think.
Many drugs, be it antihistamines, migraine medication (also prophylactic ones), dry eye medication, simple painkillers, and so on, either loose their efficacy, or produce additional, sometimes severe, side effects, or cause dependencies even when no longer helping. I have gone through that several times, it is a horrific and terrifying experience, especially as most often doctors diagnosed it wrong or denied the possibility of what I experienced. Side effects showing up after months or years are extremely hard to figure out, and things go badly wrong when it is simply diagnosed as an additional health issue without realizing what is causing it. Some long term issues are at least documented on the internet (birth control pills causing emotional issues after months or years of use), for others there are anecdotal stories to be found (Allegra no longer effective though causing bad withdrawal symptoms even when tapered off). But for others it is up to the patent to figure it out (Pazeo and Bromfenac interfered and caused scratchy eyes after an year; Tyrvaya caused nose bleed and constant nose and face pain after two years; Imitrex having scary cognitive side effects after years of somewhat regular use). Often it is up to the patient to try to get off the harmful medication without any medical support and to find alternative treatment methods not causing harm, because no studies support the personal experience.
I didn't notice any mention in the article or in the comments regarding Phase 4 trials. Such trials were envisioned as the mechanism by which longer-term effects along with questions of clinical utility and population characteristic variability could be addressed. Perhaps Sensible Medicine could put Phase 4 trials on its discussion agenda.
As Dr. Miller points out in his excellent comment below, using the figures for relative risk reduction is the major tool of deception. This is especially true for analysis of studies with low incidence of hard endpoints which is true for almost all cardiovascular disorders. The inclusion of a number of softer or subjective outcomes in order to create a primary composite endpoint is another wrinkle in the deceptive strategy. We have all seen studies where each individual outcome shows insignificant differences but, through the magic of statistical prestidigitation, the composite is deemed statistically significant. I can also recall a few recent studies reviewed in Sensible Medicine where proposed durations of say five years are terminated after two or three years because the intervention being investigated showed a two or three percent "benefit" as compared to the control. The implicit assumption there is that the benefit margin is significant and will continue to increase over time. In addition to cardiovascular diseases, practically all "preventive medicine" studies suffer from these deficiencies. The statin drugs are the poster boys for this sort of deception. In my opinion, statistical significance and more recently favored statistical surrogates such as number needed to treat and noninferiority should be disregarded. Anyone of average intelligence can look at the absolute risk reduction of clinically significant endpoints and make an individual decision on whether a recommended treatment is worthwhile.
This illustrates one more way Big Pharma controls medicine in the USA and deceives the medical community.
Another powerful tool of deception is using relative benefits instead of actual benefits. For statins, the relative benefit is a 29% reduction in death, whereas the actual (absolute) reduction in death is only 1.2%. In my case, benefit from taking statins would drop my probability of death from 38% to 37% (insignificant). Yet, Big Pharma always minimizes the side effects by using absolute numbers not relative numbers. I feel so much better getting off the statins (cramps gone, brain fog gone, strength improved). Egregious behavior that so many physicians have fallen for.
I wish the conspiracy-minded use of the 'Big Pharma' shibboleth would go away. The big corporations in banking, energy, transportation, pharmaceuticals, etc are not inherently bad or good, but in our modern world they are indispensable and mostly populated by people trying to be something useful. All those throwing stones too often stand in the shards of glass that were once lovely glass houses...
Psychiatrist here. This is a HUGE problem if we are assuming that our patients are chronically ill, with illness often manifesting in the patients' teens or twenties, and lasting in some form for their entire lifetime. (My mother-in-law just contacted me about a 90 year old friend who was just psychiatrically hospitalized with chronic depression and has frequent "reactions to medications.)" How much has her metabolism changed since the onset of her symptoms? How have her hormones changed, as one example? And since non-medication modes of treatment are also available, how have her coping skills changed? I would also point out that an eight week trial would only be roughly two menstrual cycles. Lots of room for improvement in the studies.
Hmm. Very interesting. I've been on BP and cholesterol drugs since 2008 with twelve years to go before I hit 90. My wife expects me to make it beyond that threshold!
How refreshing to hear a medical expert humbly admit, “I don’t know.”
This is one of the reasons for medicine being an art and how important it is with a long, stable and confident relationship with our patients.
There have been some studies with long term observational data to verify the findings of randomized controlled trials. Observational data typically involve a population that is not as selected with the criteria used in a trial and thus may show different results. I have not done any long-term cohort studies myself but my colleagues at the University of Alabama at Birmingham (UAB) have done several studies examining the use of various treatments:
https://www.ahajournals.org/doi/full/10.1161/STROKEAHA.124.048385
as well as examining the impact of SPRINT in the REGARDS observational population
https://academic.oup.com/ajh/article/36/9/498/7209344 “
Following people who participated in a trial for many years after the trial ends presents the practical problem that medical technology changes so that participants may switch from what they received in the trial to something that may be newer, cheaper, more expensive, or maybe more effective.
Love this observation John.
In addition to study length vs treatment duration: the patient in front of you has many variables ( age, coexisting conditions, concomitant drugs, differing lifestyles etc) from study participants that further complicate the “decision to treat “.
"Czar of evidence based practice" is very good
This clinical question is pervasive and implicates multiple medications in most patients (at least in cardiology, but I suspect in any field that deals with chronic health conditions). I saw many patients today where this would be a conundrum for many meds.
Shared decision making- let your patients know where evidence ends (as in extending the findings of clinical trials beyond their trial duration), and where (hopefully) expert opinion begins. And let them choose what to do. I have been taking a statin for > 20 years for primary prevention- there is no evidence that I can quote to support this practice. But, I know that my father had severe CAD, and had a 5 vessel CABG at a much earlier age. So I live with the uncertainty of whether long-term statins for primary prevention is good or not. I choose to believe (right or wrong) that clinical benefits continue to accumulate over time (in the case of statins to lower LDL), and thus I choose to continue to take them.
The lack of long-term followup is one large problem; thanks for the post. A related problem is that effectiveness in the real world is normally less than in the trial, for many reasons. For both of these, CMS could require pharma and device makers to carry out long-term observational studies, and should. No one else can, I think.
Bernie Black, Northwestern University
Thank you. i learned some new things from this piece. I appreciate the topic being addressed here.
Many drugs, be it antihistamines, migraine medication (also prophylactic ones), dry eye medication, simple painkillers, and so on, either loose their efficacy, or produce additional, sometimes severe, side effects, or cause dependencies even when no longer helping. I have gone through that several times, it is a horrific and terrifying experience, especially as most often doctors diagnosed it wrong or denied the possibility of what I experienced. Side effects showing up after months or years are extremely hard to figure out, and things go badly wrong when it is simply diagnosed as an additional health issue without realizing what is causing it. Some long term issues are at least documented on the internet (birth control pills causing emotional issues after months or years of use), for others there are anecdotal stories to be found (Allegra no longer effective though causing bad withdrawal symptoms even when tapered off). But for others it is up to the patent to figure it out (Pazeo and Bromfenac interfered and caused scratchy eyes after an year; Tyrvaya caused nose bleed and constant nose and face pain after two years; Imitrex having scary cognitive side effects after years of somewhat regular use). Often it is up to the patient to try to get off the harmful medication without any medical support and to find alternative treatment methods not causing harm, because no studies support the personal experience.
I didn't notice any mention in the article or in the comments regarding Phase 4 trials. Such trials were envisioned as the mechanism by which longer-term effects along with questions of clinical utility and population characteristic variability could be addressed. Perhaps Sensible Medicine could put Phase 4 trials on its discussion agenda.
As Dr. Miller points out in his excellent comment below, using the figures for relative risk reduction is the major tool of deception. This is especially true for analysis of studies with low incidence of hard endpoints which is true for almost all cardiovascular disorders. The inclusion of a number of softer or subjective outcomes in order to create a primary composite endpoint is another wrinkle in the deceptive strategy. We have all seen studies where each individual outcome shows insignificant differences but, through the magic of statistical prestidigitation, the composite is deemed statistically significant. I can also recall a few recent studies reviewed in Sensible Medicine where proposed durations of say five years are terminated after two or three years because the intervention being investigated showed a two or three percent "benefit" as compared to the control. The implicit assumption there is that the benefit margin is significant and will continue to increase over time. In addition to cardiovascular diseases, practically all "preventive medicine" studies suffer from these deficiencies. The statin drugs are the poster boys for this sort of deception. In my opinion, statistical significance and more recently favored statistical surrogates such as number needed to treat and noninferiority should be disregarded. Anyone of average intelligence can look at the absolute risk reduction of clinically significant endpoints and make an individual decision on whether a recommended treatment is worthwhile.
This illustrates one more way Big Pharma controls medicine in the USA and deceives the medical community.
Another powerful tool of deception is using relative benefits instead of actual benefits. For statins, the relative benefit is a 29% reduction in death, whereas the actual (absolute) reduction in death is only 1.2%. In my case, benefit from taking statins would drop my probability of death from 38% to 37% (insignificant). Yet, Big Pharma always minimizes the side effects by using absolute numbers not relative numbers. I feel so much better getting off the statins (cramps gone, brain fog gone, strength improved). Egregious behavior that so many physicians have fallen for.
I wish the conspiracy-minded use of the 'Big Pharma' shibboleth would go away. The big corporations in banking, energy, transportation, pharmaceuticals, etc are not inherently bad or good, but in our modern world they are indispensable and mostly populated by people trying to be something useful. All those throwing stones too often stand in the shards of glass that were once lovely glass houses...
Psychiatrist here. This is a HUGE problem if we are assuming that our patients are chronically ill, with illness often manifesting in the patients' teens or twenties, and lasting in some form for their entire lifetime. (My mother-in-law just contacted me about a 90 year old friend who was just psychiatrically hospitalized with chronic depression and has frequent "reactions to medications.)" How much has her metabolism changed since the onset of her symptoms? How have her hormones changed, as one example? And since non-medication modes of treatment are also available, how have her coping skills changed? I would also point out that an eight week trial would only be roughly two menstrual cycles. Lots of room for improvement in the studies.
Hmm. Very interesting. I've been on BP and cholesterol drugs since 2008 with twelve years to go before I hit 90. My wife expects me to make it beyond that threshold!